metadata
dataset_info:
features:
- name: code_id
dtype: string
- name: result_source
dtype: string
- name: judging_llm
dtype: string
- name: business_rule_percentage
dtype: int64
- name: format_readability_rate
dtype: int64
- name: technical_jargon_rate
dtype: int64
- name: coverage_rate
dtype: int64
- name: testability_rate
dtype: int64
- name: atomicity_rate
dtype: int64
- name: faithfulness_rate
dtype: int64
- name: error
dtype: 'null'
splits:
- name: train
num_bytes: 546
num_examples: 7
download_size: 4981
dataset_size: 546
configs:
- config_name: default
data_files:
- split: train
path: data/train-*